Generative Profiling for Soft Real-Time Systems and its Applications to Resource Allocation

2026-04-01

AI Summary

Existing approaches struggle to accurately capture the fine-grained temporal behavior of tasks under varying resource contexts, limiting resource utilization efficiency in soft real-time systems. This work proposes a generative analytical method based on nonparametric conditional multi-marginal Schrödinger bridges (MSB), introducing this framework for the first time to real-time system modeling. The method synthesizes high-fidelity task execution traces under unobserved resource configurations while providing maximum likelihood guarantees. By incorporating hardware resource context for adaptive inference, it significantly improves resource utilization efficiency in multicore real-time systems, as demonstrated through extensive evaluations on real-world benchmarks, thereby validating its effectiveness and practicality.

Abstract

Modern real-time systems require accurate characterization of task timing behavior to ensure predictable performance, particularly on complex hardware architectures. Existing methods, such as worst-case execution time analysis, often fail to capture the fine-grained timing behaviors of a task under varying resource contexts (e.g., an allocation of cache, memory bandwidth, and CPU frequency), which is necessary to achieve efficient resource utilization. In this paper, we introduce a novel generative profiling approach that synthesizes context-dependent, fine-grained timing profiles for real-time tasks, including those for unmeasured resource allocations. Our approach leverages a nonparametric, conditional multi-marginal Schrödinger Bridge (MSB) formulation to generate accurate execution profiles for unseen resource contexts, with maximum likelihood guarantees. We demonstrate the efficiency and effectiveness of our approach through real-world benchmarks, and showcase its practical utility in a representative case study of adaptive multicore resource allocation for real-time systems.

Problem

• real-time systems
• timing behavior
• resource allocation
• context-dependent profiling
• execution time analysis

Innovation

• generative profiling
• real-time systems
• multi-marginal Schrödinger Bridge
• resource allocation
• timing behavior modeling